边际回归模型方法及医学应用

Marginal Regression Models and its Medical Application

目的探讨边际回归模型在医学研究领域多变量相关分析中的应用。方法结合两个医学研究实例,说明边际回归模型原理,软件和实际应用。结果连续型和二分类的多变量需要使用不同的相关测量。对两种类型资料边际回归模型都可以同时估计回归系数和关联参数。结论边际回归模型方法应用灵活、方便,可作为多变量相关分析的常规工具之一。

Objective To explore the application of marginal regression models on multivariate correlation analysis in medical research.Methods Two medical examples illustrated marginal regression models and its application.Results Continuous and binary data need different correlation measure.Marginal regression models can be used to estimate regression coefficients and association parameters for two kinds of data.Conclusion Marginal regression models can be one of common tools in multivariate correlation analysis.